SaaS MVP is finally ready for BETA. My goodness what a ride

I've been building TWO Docs since May 2026. The first time I said it was ready, I was excited. Later I realized how freakin' flawed my SaaS really was. Inconsistent UI, all the way to bugs I didn't even know fucking existed. And after an audit with Claude Code, it only got worse.

Read also: Claude said we fixed the SaaS MVP bugs, but I don't fucking know

It was only after I started using my own docs editor every single day that I discovered how bad my MVP was. An audit tells you what's broken. Using it every day shows you how much it annoys you. Thank fuck I didn't push it to marketing yet.

Read also: I underused my own SaaS app, until now

But after four months, hundreds of iterations, and 989 deployments to be exact, the app is finally in a state where it should be somewhat presentable.

I still think the light mode is questionable, but I can fix that along the way. The point is, I stuck to the plan. I went over every component in TWO Docs, and I didn't rush any of it.

No real setbacks this time

I could explain in detail everything I found, but that would bore anyone to death. I listed most of it in a previous post, which covers everything up to the point where I got to planner, activity, library, and studio.

I was expecting real setbacks with those four. As it turns out, there weren't any. Sure, a lot of the iterations and bug fixes didn't happen in that cool one-shot prompting way Redditors like to brag about.

That's probably because every time I paste the committed result from Claude Code or Cursor into Claude, it starts rambling about what's fixed, what needs verifying, and which holes I still need to plug. Nothing gets waved through.

I understand what we're doing much better now than I did in May. That doesn't mean I can read all the technical output and pretend I'm the pro here. I'm not.

Read also: I build software like a toddler learning to walk

I think the other reason I had fewer setbacks is that I stopped shooting at everything to see what sticks. I took notes over the past two weekends and went over everything I could find.

PS: Whatever you see in the screenshots below, this is my real account, with a lot of messy test docs too and real blog posts I've written.

Planner, activity, and library

And it was a lot. UX-wise, it was a nightmare. A lot of items couldn't be renamed, deleted, or moved, and I had to think through a lot of decisions.

Planner had to stay strict.

It shouldn't become a replacement for a full task manager, but should hold tasks that are tied to documents.

Activity was missing a lot, and I had to make sure it showed the right things without feeling cluttered.

The library needed a complete revamp, because it was missing half the shit I'd added to the app.

Those were the real decisions, and I kept them strict so the SaaS stays intact and doesn't turn into Notion.

Studio was the real challenge

I added Studio a long time ago, thinking it would solve the most creative problems. Then I had to take a step back and think about what a BETA should look like.

I called it Studio because it's where ideas should live, or brew. I wanted something better than just writing ideas down in the planner.

Apart from split view, I think it's the coolest feature TWO Docs has. You can add an unlimited number of ideas, and if you decide to write about one, you don't need to go to Docs and create a document. You turn it into a doc right from that page.

The biggest challenge was syncing. If someone changes the title after converting an idea into a doc, that change has to find its way back to Studio. Now it does, so there's never a mismatch.

The moment you convert an idea, its status automatically changes from Not started to In progress. When you're done writing, you can delete the idea or mark it as Published. Which makes me think I need to change that to Done.

Published is really a blogger or writer word, and it narrows things down too much.

I think this is one of the top features I have to offer, especially because of how creatives work. You should be able to go from writing down an idea straight into writing it, without clicking three more buttons first.

The gap between having an idea and starting on it is where a lot of ideas die, and I didn't want my app making that gap any wider.

Wall and canvas were the same fucking thing

I had both a Wall and a Canvas as options. Only this week did I realize they're pretty much the same fucking thing, except Canvas is infinite with zoom and Wall is more fixed.

I don't know why my brain back then thought it was a good idea to put two almost identical options on one Studio page, but okay. I deleted Wall and went all in on Canvas.

What I'm aiming for, and still am, is a simplified version of Apple Freeform. A lot of creatives and writers I know like to start with something freestyle or visual before they write. A visual representation can do wonders.

So I started rethinking how Canvas should work. Right now you can add past docs, notes, free text, images, all sorts of shapes and color swatches.

I think that's enough for a BETA. I need feedback from real users before I make it more complicated.

I was always scared

With every change I made, big or small, I kept asking myself if it would break the docs editor or the core essence of the app. My first rounds of changes constantly broke things. Docs disappeared, text didn't save, and it was the most fucking frustrating time.

Now I think it's not that bad that it happened. After solving it and debugging what I could find, my reflex is to ask, "Will this break real docs, the editor, or live syncing?" before anything else.

That makes me more careful. I fix one thing at a time instead of trying to one-shot an entire section.

There are so many lessons in that, and they deserve their own blog post.

What's next?

So far I've only tackled private workspaces. But I know that during these weeks of debugging and fixing, Claude still included shared workspace logic to some degree.

That leaves me with two directions:

I could start rebuilding the Mac app first, or I could check the shared workspace environment and see if it's done properly. I already know it's not.

Read also: Mac app paused, recalibration, and content: update #12

Just like I added a BETA label next to Studio, I think I should add one to shared workspaces until I've gone through them.

My balance is going to change too. Developing is one thing, but it's really time to build a marketing plan, invite BETA users, and revisit what I originally planned.

My instinct says fix shared workspaces first, even though I clearly said the Mac app would be private workspaces only for now. I'll let it sink in until tomorrow, when I write my Friday close post for the week.

Read also: My MVP is ready. I still don't have a marketing plan.

What a wild ride

So my SaaS MVP is finally fucking ready for BETA. I think the dark mode and layout are beautiful. The app is minimal, functional, and doesn't come with an overload of features nobody fucking wants.

Before, I would have called it ready and believed it. I did, the first time. The difference now is that I've used it every day, broken it, fixed it, and gone through every corner of it with notes next to me. I could still be dead wrong.

But I'm confident this is the genuine shape of an MVP that's presentable, and that it's finally something I want other people to open.

989 deployments in, and the part where other people get to use it hasn't even started.

What a wild ride it was so far.